Change in ...nfs-ganesha[next]: From 2.8 onwards, ASAN not working for libganesha_nfsd
by Ashish Sangwan (GerritHub)
Ashish Sangwan has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/492079 )
Change subject: From 2.8 onwards, ASAN not working for libganesha_nfsd
......................................................................
From 2.8 onwards, ASAN not working for libganesha_nfsd
Currently getting this warning when I enabled address sanitizer:
CMake Warning at cmake/modules/sanitize-helpers.cmake:161 (message):
AddressSanitizer disabled for target ganesha_nfsd because there is no
sanitizer available for target sources.
Call Stack (most recent call first):
cmake/modules/FindASan.cmake:58 (sanitizer_add_flags)
The reason is because the library libganesha_nfsd is made up entirely
of object libraries without any source files of its own and hence
the cmake script sanitizer_add_flags rejects enabling sanitizer for it.
Current patch removes creating of FsalCore as Object library and instead
directly add the c files as sources of target ganesha_nfsd. This is inline
to what is suggessted in the cmake documentation:
"Some native build systems (such as Xcode) may not like targets that have only
object files, so consider adding at least one real source file to any target that
references $<TARGET_OBJECTS:objlib>."
https://cmake.org/cmake/help/v3.12/command/add_library.html
The other change in this patch is that when SANITIZE_ADDRESS is enabled we
should not use --no-undefined as mentioned in the ASAN docs:
Q: When I link my shared library with -fsanitize=address, it fails due to some
undefined ASan symbols (e.g. asan_init_v4)?
A: Most probably you link with -Wl,-z,defs or -Wl,--no-undefined.
These flags don't work with ASan unless you also use -shared-libasan
(which is the default mode for GCC, but not for Clang).
https://github.com/google/sanitizers/wiki/AddressSanitizer
Signed-off-by: ashish.sangwan <ashishsangwan2(a)gmail.com>
Change-Id: Ibaf837804eed595eaa990b40b1ea5e11a229d571
---
M src/MainNFSD/CMakeLists.txt
1 file changed, 9 insertions(+), 8 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/79/492079/1
--
To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/492079
To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings
Gerrit-Project: ffilz/nfs-ganesha
Gerrit-Branch: next
Gerrit-Change-Id: Ibaf837804eed595eaa990b40b1ea5e11a229d571
Gerrit-Change-Number: 492079
Gerrit-PatchSet: 1
Gerrit-Owner: Ashish Sangwan <ashishsangwan2(a)gmail.com>
Gerrit-MessageType: newchange
4 years, 7 months
Announce Push of V4-dev.17
by Frank Filz
Branch next
Tag:V4-dev.17
Merge Highlights
* Fix ganesha_stats v3_full or v4_full fails.
* stats: Adds json output to some stats command in ganesha_stats.py
* selinux: revert systemd_logind_t dbus hack
* spec - fix "installed but not packaged" error
* Report STATE_ESTALE error with LogDebug() in state_lock(), state_test()
Signed-off-by: Frank S. Filz <ffilzlnx(a)mindspring.com>
Contents:
2a62611 Frank S. Filz V4-dev.17
258bf25 Madhu Thorat Report STATE_ESTALE error with LogDebug() in
state_lock(), state_test()
80df227 Daniel Gryniewicz spec - fix "installed but not packaged" error
b38daec Kaleb S. KEITHLEY selinux: revert systemd_logind_t dbus hack
185b225 Romain Vaillant stats: Adds json output to some stats command in
ganesha_stats.py
9b8a5fc Xi Jinyu Fix ganesha_stats v3_full or v4_full fails.
4 years, 7 months
Change in ...nfs-ganesha[next]: Make NFS v3 Proxy much less noisy.
by Solomon Boulos (GerritHub)
Solomon Boulos has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/492048 )
Change subject: Make NFS v3 Proxy much less noisy.
......................................................................
Make NFS v3 Proxy much less noisy.
When writing this I didn't know about LogFullDebug. Most of my LogDebug
messages should have been LogFullDebug. I still want to log on entry for
every one of the FSAL handlers, but I don't need to be debugging the
inner loops of readdir or RPC during normal runs.
Change-Id: Ife68324d195a058513b4a1745e405b3920dcc68a
Signed-off-by: Solomon Boulos <boulos(a)google.com>
---
M src/FSAL/FSAL_PROXY_V3/main.c
M src/FSAL/FSAL_PROXY_V3/rpc.c
2 files changed, 74 insertions(+), 64 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/48/492048/1
--
To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/492048
To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings
Gerrit-Project: ffilz/nfs-ganesha
Gerrit-Branch: next
Gerrit-Change-Id: Ife68324d195a058513b4a1745e405b3920dcc68a
Gerrit-Change-Number: 492048
Gerrit-PatchSet: 1
Gerrit-Owner: Solomon Boulos <boulos(a)google.com>
Gerrit-MessageType: newchange
4 years, 7 months
Change in ...nfs-ganesha[next]: stats: Adds json output to some stats command in ganesha_stats.py
by Vaillant Romain (GerritHub)
Vaillant Romain has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/491899 )
Change subject: stats: Adds json output to some stats command in ganesha_stats.py
......................................................................
stats: Adds json output to some stats command in ganesha_stats.py
This adds the support of the json format for most commands by adding
a new method to all report objects.
This is done in two step, one method is generating a dict and the
other simply dump it in the json format. In the future the text format
could also leverage the report method, but it was kept separate to kept
changes at their minimal.
Signed-off-by: Romain Vaillant <romain.vaillant(a)scality.com>
Change-Id: I849b751c6735847e8152713bf2807c656dac0ea3
---
M src/scripts/ganeshactl/Ganesha/glib_dbus_stats.py
M src/scripts/ganeshactl/ganesha_stats.py
2 files changed, 662 insertions(+), 93 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/99/491899/1
--
To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/491899
To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings
Gerrit-Project: ffilz/nfs-ganesha
Gerrit-Branch: next
Gerrit-Change-Id: I849b751c6735847e8152713bf2807c656dac0ea3
Gerrit-Change-Number: 491899
Gerrit-PatchSet: 1
Gerrit-Owner: Vaillant Romain <git(a)rphv.dev>
Gerrit-MessageType: newchange
4 years, 7 months
Change in ...nfs-ganesha[next]: selinux: revert systemd_logind_t dbus hack
by Kaleb KEITHLEY (GerritHub)
Kaleb KEITHLEY has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/491553 )
Change subject: selinux: revert systemd_logind_t dbus hack
......................................................................
selinux: revert systemd_logind_t dbus hack
it doesn't build in rhel8
and shame on me for believing I understood selinux well enough to
trust that it would work
Change-Id: I518dd48e3c68327b035e26d7da9f5f31a39865cf
Signed-off-by: Kaleb S. KEITHLEY <kkeithle(a)redhat.com>
---
M src/selinux/ganesha.te
1 file changed, 0 insertions(+), 11 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/53/491553/1
--
To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/491553
To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings
Gerrit-Project: ffilz/nfs-ganesha
Gerrit-Branch: next
Gerrit-Change-Id: I518dd48e3c68327b035e26d7da9f5f31a39865cf
Gerrit-Change-Number: 491553
Gerrit-PatchSet: 1
Gerrit-Owner: Kaleb KEITHLEY <kaleb(a)redhat.com>
Gerrit-MessageType: newchange
4 years, 7 months
Change in ...nfs-ganesha[next]: MDCACHE - Fix cache trust when file is modified during getattr
by Name of user not set (GerritHub)
yoni(a)weka.io has uploaded this change for review. ( https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/491544 )
Change subject: MDCACHE - Fix cache trust when file is modified during getattr
......................................................................
MDCACHE - Fix cache trust when file is modified during getattr
When mdcache_refresh_attrs is called, first it refreshes the attributes and
then it sets MDCACHE_TRUST_ATTRS so the next getattr calls will use the
cached results. If write happened while attrs are being refreshed it means
the attrs are stale at that point and shouldn't be cached.
Introducing attr_generation solves that problem by increasing the generation
every time a write occurs, comparing the generation before and after
refreshing attributes and setting the MDCACHE_TRUST_ATTRS flag only if the
generation didn't change during that time.
Change-Id: I9f22aac2046b305c4d3796e05526781eacbbbe7d
Signed-off-by: Yoni Knobil <yoni(a)weka.io>
---
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_file.c
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_handle.c
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_int.h
M src/FSAL/Stackable_FSALs/FSAL_MDCACHE/mdcache_lru.c
4 files changed, 19 insertions(+), 2 deletions(-)
git pull ssh://review.gerrithub.io:29418/ffilz/nfs-ganesha refs/changes/44/491544/1
--
To view, visit https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/491544
To unsubscribe, or for help writing mail filters, visit https://review.gerrithub.io/settings
Gerrit-Project: ffilz/nfs-ganesha
Gerrit-Branch: next
Gerrit-Change-Id: I9f22aac2046b305c4d3796e05526781eacbbbe7d
Gerrit-Change-Number: 491544
Gerrit-PatchSet: 1
Gerrit-Owner: yoni(a)weka.io
Gerrit-MessageType: newchange
4 years, 7 months
Announce Push of V4-dev.16
by Frank Filz
I missed some patches on Friday and Daniel needs the selinux patch to
backport into V2.8.4, so here we go with a quick merge.
Branch next
Tag:V4-dev.16
Merge Highlights
* FSAL: add ERR_FSAL_STILL_IN_USE
* FSAL_UP: add new "try_release" upcall
* FSAL_CEPH: set ino_release_cb on mount
* selinux: additional policy
Signed-off-by: Frank S. Filz <ffilzlnx(a)mindspring.com>
Contents:
e6e65d4 Frank S. Filz V4-dev.16
bb0d5a0 Kaleb S. KEITHLEY selinux: additional policy
382966c Jeff Layton FSAL_CEPH: set ino_release_cb on mount
58a9114 Jeff Layton FSAL_UP: add new "try_release" upcall
c36c553 Jeff Layton FSAL: add ERR_FSAL_STILL_IN_USE
4 years, 7 months
RHEL clients randomly missing to send RECLAIM_COMPLETE op and ganesha sends NFS4ERR_GRACE indefinitely
by des@vmware.com
In our system test setups with NFSv4.1 mounts, we keep hitting this RHEL client bug where under some unknown condition client fails to send RECLAIM_COMPLETE_OP after client id expiry and recovery. And ganesha keeps sending NFS4ERR_GRACE indefinitely and this client will hit timeout on all the OPENs and hence halting the system testing.
We upgraded to RHEL7.7 but still issue persists. I tried searching in redhat bugs but do not see any outstanding issues about this. Is anyone else here facing this issue? Do we know if this is really a RHEL client issue?
<ganesha.log>
2020-04-18T16:28:00Z : epoch 5e983c8e : w1hs3e1512.vsanstfsad.local : ganesha.nfsd-164[::ffff:172.30.63.199] [svc_495] 1223 :nfs4_op_open :STATE :Entering NFS v4 OPEN handler -----------------------------
2020-04-18T16:28:00Z : epoch 5e983c8e : w1hs3e1512.vsanstfsad.local : ganesha.nfsd-164[::ffff:172.30.63.199] [svc_495] 1283 :nfs4_op_open :STATE :OPEN Client id = 5e983c8e00000026
2020-04-18T16:28:00Z : epoch 5e983c8e : w1hs3e1512.vsanstfsad.local : ganesha.nfsd-164[::ffff:172.30.63.199] [svc_495] 507 :create_nfs4_owner :STATE :Previously known owner {STATE_OPEN_OWNER_NFSV4 0x7fe1475ff610: clientid={0x7fe177711ab0 ClientID={Epoch=0x5e983c8e Counter=0x00000026} CONFIRMED Client={0x7fe1777119e0 name=(45:Linux NFSv4.1 w1h34v25-c0017.vsanstfsad.local) refcount=1} t_delta=0 reservations=2 refcount=4} owner=(24:0x6f70656e2069643a0000002c00000000000000c4d78a2b49) confirmed=1 seqid=0 refcount=2} is being reused
2020-04-18T16:28:00Z : epoch 5e983c8e : w1hs3e1512.vsanstfsad.local : ganesha.nfsd-164[::ffff:172.30.63.199] [svc_495] 1322 :nfs4_op_open :NFS4 :open4_validate_claim failed
2020-04-18T16:28:00Z : epoch 5e983c8e : w1hs3e1512.vsanstfsad.local : ganesha.nfsd-164[::ffff:172.30.63.199] [svc_495] 1409 :nfs4_op_open :STATE :failed with status NFS4ERR_GRACE
2020-04-18T16:28:00Z : epoch 5e983c8e : w1hs3e1512.vsanstfsad.local : ganesha.nfsd-164[::ffff:172.30.63.199] [svc_495] 983 :nfs4_Compound :NFS4 :Status of OP_OPEN in position 2 = NFS4ERR_GRACE, op response size is 4 total response size is 92
2020-04-18T16:28:00Z : epoch 5e983c8e : w1hs3e1512.vsanstfsad.local : ganesha.nfsd-164[::ffff:172.30.63.199] [svc_495] 1112 :nfs4_Compound :NFS4 :End status = NFS4ERR_GRACE lastindex = 2
<ganesha.log>
We had discussed this before in this community and ganesha behaves as per RFC, so we dint do anything:
https://lists.nfs-ganesha.org/archives/list/devel@lists.nfs-ganesha.org/m...
4 years, 7 months